ion source brightnesswhite
TRANSCRIPT
Plasma Ion Source Brightness and Efficiency
Nick White Albion Beams Inc.
Computer Modeling used: • SORCERY: My own program from 1984
– Based on Sheffield program by Dirmikis, 1970s – Added plasma modeling for ion source extraction – Added random momentum at emitting surface (with restrictions) – Now can be run in DOSbox (freeware) (2013)
• OPERA suite, TOSCA, SCALA – $35000 plus annual fees plus extras – Originated Rutherford Laboratory, 1970s – Commercialized, now owned by Cobham – Extremely powerful highly accurate magnet modeling – Links to 3D CAD files – 3D finite element Poisson solver including
• Space-charge, with multiple beams, sources, secondaries • Plasmas (with some limits) • Non-linear magnetics • RF, hysteresis, motors, forces, induction…
Emittance and Brightness
• The use of normalized 2D emittance – Transverse momentum scales with angle * sqrt(energy) , mass being constant
• Usually mass is omitted, energy is in MeV
• I have recently worked with beams at energies down to 200eV – So transverse momentum becomes a sizeable fraction of the total momentum – In this presentation I mainly talk about the narrow dimension
• Brightness is current / 4D emittance – BUT I work with ribbon-shaped beams (mostly), – For a ribbon beam, you can’t simply square the 2D emittance
Positive source/Charge exchange canal combination for negative beams
A Hot Cathode Penning Arc Source (Bernas with anticathode)
Indirectly Heated Cathode (500 hour service life)
Applied 200-300 Gauss B-field
Anticathode
Typical positive ion beam extraction from a plasma source
From Forrester
Cutaway of divergent ribbon beam source
Plasma 20,018V
Beam
+20,000V
-5kV
0V
10keV 60 mA BF3 gas best case beam Ribbon beam, cross section, 100mm in long direction
3mm wide 4mm wide 5mm wide
10keV 60 mA B best case beam ε= 6.8 mm. mrad.(MeV)1/2
Of the total current, 12mA is B+, remainder is BF+, F+, BF2+,B++
We believe these figures because the total divergence is measured, but real, quantitative emittance measurements are rare.
Degrees
Long transverse dimension of ribbon beam, mm
10keV 60 mA B best case beam Emittance in long direction: Ends of the slot cause locally ugly emittance 90% lies within parallelogram 100mm x 10mrad with designed divergence ε= 100 mm.mrad.(MeV)1/2 therefore average B=1.76 µA/(mm2.mrad2.MeV) Peak brightness much higher but optical aberrations dominate.
Degrees
Real photo of the 200mm tall beam discussed in the following slides
• Beam: Boron at 200eV to 5 keV, 2.5 mA • Detector: Fax paper placed 150mm
beyond an aluminum plate with holes in it.
Aperture plate
Beam burns 150mm beyond plate
Measured emittance
Nick White, John Chen, Chris Mulcahy, Sukanta Biswas, Russ Gwilliam , IIT 2006
Beam Divergence and Angle Control
Insignificant change in overall angles down to 200eV, but each beamlet spreads more at low energy.
Horizontal Angular spreadBoron 5 keV
-4
-3
-2
-1
0
1
2
3
4
-30 -20 -10 0 10 20 30
Horizontal distance, mm
Hor
izon
tal a
ngle
, deg
rees
Horizontal Angular spreadBoron, 200eV
-4
-3
-2
-1
0
1
2
3
4
-30 -20 -10 0 10 20 30
Horizontal distance, mm
Hor
izon
tal a
ngle
, deg
rees
Angular spread comes from thermal physics
• Lateral motion of beam ions in a beamlet is +/- 1680m/s max, or ~1160 m/s rms
• Ion source gas temperature is ~ 1500K – Wall temperature measured at 1000K
• From the thermal motion of the ion source gas, the transverse rms velocity of the ions is > 700 m/s
• The observed transverse momentum is close to that of the source gas. It is thermal.
41862 m/s (for 200eV B)
+/- 1.15 degrees (measured)
+/-840 m/s (for 95%)
Physics of beam formation
• Generation of ions within the plasma – Electron energy distribution and cross sections
• Formation of plasma in chamber – Issues of thermalizing electrons – Potential distribution and temperature
• How the ions leave the plasma • The plasma sheath • The extraction field • The beam plasma
Generation of equilibrium plasma • σi peaks between 60 and 100 eV
• Electrons from a hot cathode at -60 V – Magnetic field blocks electrons from anode
• Gas molecules at 0.03eV -> thermal ions – Unless we dissociate a molecule
• One fast electron -> 2 slow electrons
• Elastic scattering cross section rises as electrons slow down -> electron diffusion to anode rises as electrons slow down (paradox)
– Mid-energy electrons lost to the walls – Cooler electrons diffuse across magnetic field – A cool (Te~4eV) plasma is generated – Plasma potential reaches equilibrium at ~20V w.r.t. anode – Thus the primary electron energy in most of the plasma is ~ 80eV
Electron distribution/Potential distribution
Ne.v = Je
Plasma
Sheath
0.5kT
4 to 5 kT
To cathode
Anode
Distance across arc chamber
Plasma electrons
Cathode electrons
Ion energy spectrum • Ion energy comes from:
– Initial gas temperature (~0.1 eV) • random
– Acceleration across a part of the 2V potential variation within the plasma • On average this will be ~ 1.5 eV in beam direction
– Acceleration across the plasma sheath • Depending on species, ~ 18eV in beam direction
• After leaving the box, ions gain energy from the full extraction voltage
• The distribution peaks at around 19.5V – low-energy tail (ions created in or near the sheath)
• Axial FWHM of order kTe
– Transverse FWHM is still thermal
Extraction region • The high electric field of the extraction region is terminated on the plasma • The emission of ions from the plasma is space-charge limited (Child-Langmuir law) • The delivery of the ions to the sheath is solely determined by plasma physics • If current is too high, plasma sheath pushes out of the box toward the high voltage,
raising the electric field until it can extract all the ions from the plasma • If current is too low, the electric field pushes the plasma away from the exit aperture • Good beam formation requires a very fine balance
Source plasma
Beam plasma
The Beam Plasma
• At right hand side, electrons knocked of walls or from beam collisions are sucked into the beam.
• We use accel/decel beam extraction to block these from being sucked out of the beam by the positive ion source.
• They reach an equilibrium temperature of a few eV, and limit electric fields within the beam.
J= 109
J=119
J=112
Dependence of arsenic beam divergence on Gap
Modeled in SCALA
Gap in mm
Degrees
Plasma
Extraction electric field Sheath (dynamic)
Defined Plasma
Pre- sheath
Sheath
Beam
1V equipotentials
Beam
Defined plasma
Orange: surface of dynamically solved plasma
Last night I ran a 32-year old model in SORCERY on my laptop, using the new DOSbox freeware.
From 1981 Doctoral Thesis
Conclusions
• For maximum brightness, a high density cool plasma in the ion source is needed
• Transverse emittance is very low • Axial energy spread depends more on plasma
temperature and noise, but is still low • Beam emittance is dominated by defects in
beam formation and focusing
Efficiency • As the ionization of the gas or vapor increases,
there will be changes in beam composition: – More doubly, triply-charged ions – Less surviving molecular fragments
• Clearly more ionizing electrons will raise the efficiency – For a given output, increasingly precise control of
vapor or gas flow is needed – Efficiencies of >>50% are possible
• But ions leave the plasma isotropically – Therefore >95% hit the wall
• But for many species these are recycled
Ribbon beam, with thermal motion • Arsenic, 50mA 30 keV • Beam is 100mm in the orthogonal direction
– So 0.5 mA/mm
Degrees
Extraction electrodes • Top view with ¼ of electrode cut away to show
the beam shape • Gap is 17mm • Extraction current is 65 mA • Species is As+
• Slot in source is 3mm x 100mm • Slots in suppression and ground electrodes are
4mm and 5mm • Although the beam is very well formed, it
passes very close to the slot sides, and 10% fluctuation in current will cause major beam strike
12mm wide slot in suppression and extraction
• Top view with ¼ of electrode cut away to show the beam shape
• Ion source slot is 3mm wide • Gap is 13.3mm • Extraction current is 65 mA • Species is As+
Comparison of 3 source types
Sputter (positive or negative) Penning arc discharge ECR Microwave
Energy distribution - axialSurface binding energy, Large high-energy tail
Thermal plus plasma potential distribution
Thermal plus plasma potential distribution
Energy distribution - transverseSurface binding energy, Large high-energy tail Thermal Thermal plus instability fields
Emitting areaSample size or sputtering beam size Optics of aperture - 2-6 mm Optics of aperture - 2-6 mm
Beam Formation from Source
• Modeled with my own 2D program SORCERY for a cylindrical symmetry beam • Current 8mA, mass 20, energy 40 keV, Suppression 5 kV • Gap 10mm • This model is for extraction from a true plasma.
Plasma Beam
Emittance of beam
• Emittance 1.5mm x 20 mrad max (90% within 4 mrad) • In this case, no thermal motion was assumed.
+20mrad
0mrad
-20mrad